I went to an extended lunch seminar earlier this week about "competency based recruiting". The focus of the seminar was on using competency based methods for recruiting ph.d. students in an academic context. This was the second of three lectures in a series about this topic and the speaker was Maln Lindelöw of Lindelöw & Partners. Two of my colleagues went to the first seminar and thought it was great.
The starting point is that many traditional tools for selecting a person for a job - such as documents of different kinds (CV, cover letters etc.) or job interviews - aren't that good. There's actually scientific research supporting that statement. Years of education or years of work experience say very little about future work performance. The best methods are instead structured and/or competency based interviews, work samples and ability tests.
The seminar was partly a brainstorming exercise for coming up with methods for eliciting suitable work samples in an academic context and for building cheating-proofed recruitin processes. The first step in the process though is to determine what skills and (personal, social, intellectual, leadership) abilities are most important for the job/position in question.
I was selected as a ph.d. student partly based on my bacholor's/master's thesis. A better idea would be to ask each applicant to top the thesis off by also submitting a 2-3 pages long letter where they judge the merits and the weaknesses of their own thesis. Perhaps their reflections about what they would have done different had they started to write that thesis today (i.e. lessons learned) are just as interesting as the thesis itself?
Let's say there is space for a ph.d. student in a research project that has just started. Why not send over the original research grant application and ask the applicant to (in a limited amount of time) write up a specification for the master's thesis they themselves would have liked to have done in that project. That would test many things at the same time, including what their own skills and interests are. It's also a good task as it doesn't test things they will learn in the future (as a ph.d. student), but rather things they should have learned as a master's student. You could even add another layer onto the task by setting up a 15-minute appointment (say a Skype meeting) where you give some feedback and suggestions on their master's thesis specification and ask them to (in a limited amount of time) send back an updated version after they have received feedback on their first attempt. That could also weed out having someone "help you out" with the original specification a little too much.
The thing about competency based recruitment is to think a little deeper about what the prospective ph.d will do and about the skills and abilities that are important in this project (or at this department or for you as an advisor) and then think about ways to test these important skills and abilities. Less emphasis can be placed on formal documents or references - and every recruitment process could (or should?) be customized (and different).
Perhaps best of all was that my colleagues Mario and Eva-Lotta also were at the same seminar. That means I have sparring partners to bounce ideas with should I go from theory to practice and need to recruit a ph.d. student.